About
A research led by Trevor Asserson from Asserson Law Firm and Dr. Haran Shani-Narkiss from Research for Impartial Media (RIME) examined the BBC’s portrayal of the Israel-Gaza conflict that started on October 7th 2023. Their research employed innovative methods using ChatGPT-4 to evaluate and analyze the BBC’s coverage of the war for bias, focusing on whether sympathy is expressed equally for both sides. The study examined articles’ main text and headlines, comparing BBC English and Arabic reporting, and extended the analysis to podcasts, radio, and TV.
Challenge
To validate the comprehensive datasets manually collected by the Asserson team, the researchers aimed to replicate their findings using an automated data-collection solution. By rigorously capturing all relevant publicly available data, this approach mitigates the risk of both explicit and implicit cherry-picking. To prevent any manual methods, they have used Bright Data’s Web Scraper API and were able to collect all relevant articles spanning from October 7th. The data collected and validated by the researchers, have assisted in proving their methods in identifying bias using ChatGPT-4’s large language model on online media and news agencies.
Impact
Through this collaboration, the study achieved critical visibility, influencing global media discussions about bias and transparency. Since its publication, this research has been mentioned more than 0.5 billion times in the global social media and has been widely acknowledged by outlets such as Sky News, the Telegraph, Fox News, the New York Post, the Daily mail and more. In addition, the research has secured a method using LLM to uncover bias that could potentially be used in any future studies and share a transparent perspective on various media sources around the world.
